本研究では，携帯電話位置情報データがもつ，高精度かつ膨大な時間情報に着目して，新しい需要予測手法の開発を実施した．3年間の研究成果として，おおきく以下の2点があげられる：1つ目は，膨大な時間情報があるが旅行目的などの人の意思の情報を持たない，携帯電話位置情報を解析するための方法を開発した．主には，時間変動パターンの分解手法と複数データの融合手法があれがれる．2つ目は，新幹線開業効果に着目して，その影響を従来より詳細に明らかにしたことである．とくに，従来主に考えられてきた「時間短縮効果」とは大きく異なるパターンで旅行行動が変わることを発見した．
In this research, we developed a new travel demand forecasting model, focusing on the mobile phone location data which have rich time-series information. The main achievements of this three-year research projects are following two points:
First, we have developed several methods for analyzing mobile phone location data, which has a huge amount of time information but does not have the thinking information such as the purpose of travel. They are mainly following two approaches: decomposition of time-series travel patterns and data-fusion approach for estimating the travel purpose information.
The second is the effect of new High-Speed Rail service on travel behavior has been analyzed in more detail. In particular, we found that travel behavior changes in a pattern that is significantly different from the time-saving effect that has been considered mainly in the past.
ビックデータの時間情報が拓く，新たな都市間旅行需要予測手法の開発
.png)
SC-PR-YAMAGUCHI-H-kaken 2020-9p.pdf
(707.84KB)
